Identification of Differential Gene Groups From Single-Cell Transcriptomes Using Network Entropy.

Area of Science:

• Computational Biology
• Genomics
• Bioinformatics

Background:

• Single-cell RNA sequencing (scRNA-seq) reveals cellular heterogeneity by analyzing transcriptomes.
• Identifying cell-type-specific transcriptional differences is crucial but challenging due to technical noise.
• Existing methods struggle to distinguish biological variation from noise in scRNA-seq data.

Purpose of the Study:

• To introduce DiffGE, a novel computational method for differential gene expression analysis in scRNA-seq data.
• To effectively identify highly differential gene groups by measuring expression dynamics using network entropy.
• To enhance the discovery of biologically significant transcriptional variations among cell types.

Main Methods:

• Developed DiffGE, a computational approach utilizing network entropy for gene group expression dynamics.
• Applied DiffGE to three independent scRNA-seq datasets to detect differential gene groups.
• Compared DiffGE's performance against three established differential analysis algorithms.

Main Results:

• DiffGE successfully identified highly dynamic gene groups with distinct expression patterns across cell types.
• The method demonstrated superior performance in distinguishing biological signals from technical noise.
• Gene function analysis confirmed the detected groups' significant association with cellular regulation.

Conclusions:

• DiffGE is a powerful tool for analyzing transcriptional dynamics in scRNA-seq data.
• The method effectively identifies biologically relevant differential gene groups among distinct cell types.
• DiffGE aids in understanding the mechanisms underlying cellular heterogeneity and regulation.